Environmental sustainability of prefabricated modular res ...energy embodied in the building itself (Monahan and Powell in 2001), and some have shown that operating energy is the most

Hebert, L. DOI: http://dx.doi.org/10.4225/50/58107d17d3fc5 HealthyHousing2016: Proceedings of the 7

th International Conference on Energy and Environment of

Residential Buildings, November 2016, edited by Miller, W., Susilawati, C. and Manley, K. Brisbane: Queensland University of Technology, Australia. DOI: http://dx.doi.org/10.4225/50/58107c8eb9c71

Environmental sustainability of prefabricated modular res-idential buildings compared to traditional equivalent: Two

case studies in Perth, Australia.

Prof. Peter Newman, CUSP, Australia, [email protected]

Mrs Jemma Green, CUSP, Australia, [email protected]

Mr. Roberto Minunno, CUSP, Australia, [email protected]

Ms. Francesca Geromino, CUSP, Australia, [email protected]

Abstract

The share of prefabricated modular residential buildings in the Australian construction market is

growing mainly because they are quicker to erect on-site than traditional construction, and often

cheaper; but how about their carbon footprint and more particularly their thermal performance?

To bring some light on this question, this paper uses two case studies where existing prefabricat-

ed modular buildings (a detached house and a multi-storey residential building) are compared to

their equivalent in traditional on-site construction methods. The thermal performance is assessed

using 3D modelling and energy simulation, while carbon footprint is assessed by Life Cycle Analy-

sis.

The thermal simulations showed that these prefabricated modular buildings perform better in

winter but the study also shows that low inertia modular buildings can require more cooling energy

in the summer in certain climate zones. Looking at the overall carbon footprint of the construction

elements, the LCA shows that prefabricated buildings are not always less carbon intensive than

their traditional equivalent and a wise choice of materials remains a necessity.

The discussion emphasises the importance of the local context and particularly the role of thermal

inertia in summer for Mediterranean Climate. It also demonstrates the importance of operating

energy in the overall carbon footprint of a building and, more generally, how prefabrication can

easily achieve a lower carbon footprint than on-site construction.

Keywords: Prefabricated, Modular, Carbon Footprint, Thermal Performance, Life Cycle Analysis,

Energy Efficiency

Space for a portrait

of the presenting

author

Lio Hebert

PhD Research Fellow

Curtin University Sustainability Policy Institute (CUSP)

Australia

[email protected]

mailto:[email protected]

mailto:[email protected]

Hebert, L Environmental sustainability of prefabricated modular residen-tial buildings compared to traditional equivalent: Two case studies in Perth, Australia.

HealthyHousing2016: 20-24 November, 2016, Queensland University of Technology, Brisbane, Australia.

1. Introduction

Globally, the construction industry is an important consumer of resources and producer of waste

(40% of global resources and 40% of global energy according to the 2009 report from the United

Nations Environment Programme), while at the same time, construction remains a necessity that

grows inevitably with the population growth. The materials and the technology exist to reduce the

environmental impact of buildings and early adopters such as Josh Byrne (Byrne, Eon & Newman

2014) have shown that building eco-friendly houses with minimal extra capital investment is pos-

sible, but this is not mainstream anywhere in the world and particularly not in Australia.

One solution to mainstream this performance lies in the industrialised construction method, also

called off-site construction or simply prefabrication, which has been recognised a leaner and more

efficient method of construction (D. Krug, 2013).

This paper proposes to analyse the differences in operating energy and carbon footprint of two

existing prefabricated modular residential buildings compared to Australian standard-practice

equivalent using traditional on-site construction method. More specifically, a detached house and

a multi-storey apartment block.

The literature review shows that a few researchers have already addressed the carbon and the

energy embodied in the building itself (Monahan and Powell in 2001), and some have shown that

operating energy is the most important part of the carbon footprint of a building (Aye et al. 2012).

The approach of comparing a prefabricated modular building to a traditional building to assess the

carbon footprint difference was found twice in the literature: A study in China (Cao et al. 2015)

where the two buildings were of similar typologies but not equivalent; and in the USA (Kim 2008)

where the modular house was compared to a traditional equivalent but of very similar specifica-

tions. This latter study focused on the difference in the carbon footprint of the construction process

rather than on the building itself.

2. Background of the two case studies

The first case study looks at the Stella B17 building of the ADARA project designed and manufac-

tured by the Hickory Group. It was one of the first prefabricated modular multi-storey residential

buildings in Western Australia. The building is made of fully finished volumetric modules manufac-

ture in its established Melbourne factory, and shipped to Perth. These modules are equipped with

double-glazed windows and walls and floors are made up of steel frame. The floors also received

a steel sheet and a plywood board. The building is also constituted of a concrete structure built

on-site: the ground level car park and first level floor, as well as two cores integrating the lift shafts

and staircases. The 96 modules were then piled up by crane to make 77 units on six floors. After

this the façade elements were added to dress up this elegant building.

The second case study concerns a group of 22 houses from BGC Residential, prefabricated in its

factory in Canning Vale (WA) and erected in Banksia Grove, a suburb North of Perth-metro. BGC

Residential belongs to the BGC Group, a major player in the Australian construction industry, who

decided a few years ago to start a pilot program to build prefabricated houses. The houses are

made up of two or three fully finished modules made of steel frame walls and roof structure, and

concrete slab for the floor. They were transported to site and delivered by truck without the need

of a crane. The garage and the alfresco are the only elements of the house constructed on-site.

The 22 houses are an arrangement of 2 or 3 modules making 7 concepts with different facade

colours and orientations.



3. Methodology

3.1 Carbon footprint

In our case, the carbon footprint will be evaluated by calculating the GreenHouse Gases (GHG)

emitted during the life of the building, from the extraction of its raw materials down to its decon-

struction and transport to the landfill or recycling plant.

Life Cycle Analysis (LCA) is the method chosen here to estimate the amount of GHG of our modu-

lar buildings. The unit of measurement used in our LCA is the kilogram of CO2 equivalent per

occupant per year of the life of the building, expressed as kgCO2eq/occ/y. This measures the

Global Warming Potential (GWP) (US EPA 2013) of various impacts, and means the volume of

GHG emitted by the building throughout its life.

The tool used to model the LCA was ETool, a software compliant to ISO14040 2006, ISO14044

2006, BS EN 15978 2011 standards.

For both the multi-storey building and the detached houses, all household energy consumption –

beside heating and cooling – and all water consumption were assumed to be the same between

the modular buildings and the traditional equivalents.

3.2 Operating Energy

The research being based on the comparison between two equivalent buildings from two different

construction methods, the thermal performance of the buildings envelops – and consequently the

heating and cooling loads - are to be the most significant elements to study. Therefore, it was

decided to give particular attention to these elements.

The buildings were first digitized as 3D models using Building Information Modelling (BIM) on

CYPECAD MEP (from CYPE software. CYPECAD MEP then calculates the yearly heating and

cooling demand, by hourly steps, using the EnergyPlus engine, a program developed by the U.S.

Department of Energy. This is typically expressed as a function of energy per metre square

(kWh/m2).

Weather conditions, from ASHRAE weather files, were factored into the thermal performance

modelling to consider how the building will perform in a particular climatic scenario. Additionally,

Internal comfort temperatures (set points) have been set to 21°C in winter and 24°C in summer.

The detailed load allowed to study the thermal behaviour of the building and extract important

information to understand the yearly results. The yearly results – heating and cooling loads - were

then introduced in eTool to estimate the GHG of this part of the operating energy of the buildings.

3.3 The traditional equivalent

For both the LCA and the Energy Simulation, the traditional equivalent was modelled using the

specifications given by the respective builders which would correspond to standard-practice aim-

ing at compliance. The equivalent of Hickory apartment block was modelled using concrete exter-

nal walls without insulation, brick internal walls and single-glazed windows. The BGC modular

house equivalents were modelled using double brick cavity external walls, single brick internal

walls, thicker on-site poured concrete slab and terra cotta roof tiles on timber structure.

The windows specification did not change (single glazed). The air tightness of the modular and

traditional buildings was considered the same.

This base methodology was used in the two case studies but because they happened at different

times, the parameters and assumptions were not exactly the same.



3.4 Specificities of the Multi-storey building

The Hickory building, Stella B17, was the first one to be assessed with this method and it was

modelled with all the expected internal loads (from the occupants’ bodies, the lighting and appli-

ances) to simulate a building in occupation. In eTool, the occupancy calculator estimated an aver-

age occupancy of 124.45 persons in the whole building, and a design life of 150 years for both the

modular building and the traditional equivalent.

3.5 Specificities of the detached houses

The BGC project came later and we decided that, for the purpose of a comparison, the houses

could be considered without occupancy, thus modelled and simulated with no internal loads.

It was also decided to model and simulate the 22 modular houses for Thermal Performance (TP)

but to take only best and worst performing modular houses to model their traditional equivalent for

the comparison of TP. Finally, the best performing modular house was chosen to be compared

with its traditional equivalent for the assessment of carbon footprint. In eTool, the occupancy cal-

culator estimated an average occupancy of 2.83 persons per house and a design life of 70 years

for both the modular house and the traditional equivalent.

4. Results and discussion

4.1 Multi-storey residential building

4.1.1 Heating and Cooling loads

The results for Perth Weather are represented in Table 1 below:

Table 1: Comparison of Thermal Performance of Stella B17 in Perth weather

Unexpectedly, the modular construction is not performing as well as the traditional one despite the

thermal insulation in the external walls and the double-glazed windows. These elements assist the

thermal performance in winter but something is missing in summer. To understand what can ex-

plain this difference in cooling load, the same two buildings were simulated in different climate

than the Class 4 “warm temperate” of Perth (Building Code of Australia (BCA)). They were simu-

lated in Port Hedland, Class 1 “Hot humid summer, warm winter” and in Melbourne, Class 6 “Mild

temperate”. The results are represented in Table 2 and Table 3 below:

MODULAR TRADITIONAL

[MWh] [kWh/m2] [MWh] [kWh/m

2]

Heating 26.59 4.94 39.78 7.39

Cooling 340.59 63.28 268.20 55.44

TOTAL 68.22 62.83



Table 2: Thermal performance of Stella B17 in Port Hedland compared to Perth

MODULAR PERTH PORT HEDLAND


2]

Heating 26.59 4.94 0 0

Cooling 340.59 63.28 1113.22 206.83

TOTAL 68.22 206.83

TRADITIONAL PERTH PORT HEDLAND


2]

Heating 39.78 7.39 0 0

Cooling 298.39 55.44 1175.65 218.43

TOTAL 62.83 218.43

Table 3: Thermal performance of Stella B17 in Melbourne compared to Perth

MODULAR PERTH MELBOURNE


2]

Heating 26.59 4.94 182.84 33.97

Cooling 340.59 63.28 119.16 22.14

TOTAL 68.22 56.11

TRADITIONAL PERTH MELBOURNE


2]

Heating 39.78 7.39 268.20 49.83

Cooling 298.39 55.44 83.53 15.52

TOTAL 62.83 65.35

In both cases, the modular version performs better than the traditional. Heating being not required

in Port Hedland the result is not significant for the comparison, but this time the cooling load is

lower for the modular building. On the contrary, the thermal performance in Melbourne follows the

same trend as in Perth (modular performing better than traditional in winter but worse in summer)

even if the overall performance is, this time, to the advantage of the modular building.

These results lead us to analyse the particularity of the summer conditions in these three locations,

and the two major differences were the humidity and the variance between night and day tem-

peratures. The latter made us think that the lack of thermal inertia in the modular building com-

pared to the all-concrete (walls and floors) traditional equivalent didn’t allow to make the most of

the lower nocturnal temperatures in Perth and Melbourne.

This hypothesis was verified by simulating internal walls made of bricks and finding 8% reduction

in cooling load. The approach of using bricks in the modular building to increase thermal mass is

technically feasible but not workable in terms of the construction process, cost and transport. The

second factor that influences cooling energy demand is solar gain, which is the amount of heat

that penetrates inside the building through the windows. Reducing solar gain can easily be

achieved by tinting the windows as shown in Table 4 below. In summer, solar rays penetrate the

building the most when the sun is the lowest in the sky, at sunrise and sunset, through the east

facing and the west facing sides of the building.



Table 4: Thermal Performance of the modular building with East and West windows tinted

MODULAR MODULAR

TEST 4 TRADITIONAL


2] [MWh] [kWh/m

2]

Heating 26.59 4.94 33.37 6.20 39.78 7.39

Cooling 340.59 63.28 280.36 52.09 298.39 55.44

TOTAL 68.22 58.29 62.83

This low cost and easy fix solution improved drastically the performance of the building in summer

though it reduced slightly its performance in winter.

4.1.2 Carbon footprint

The Figure 1 below shows the carbon footprint, expressed as Global Warming Potential (GWP) of

the modular building compared to the traditional equivalent:

The GHG from the building elements (foundations, envelop, fit-out, appliances, building services…)

happen only once, at the construction stage, and their carbon footprint is broken down through-out

the life of the building. However, the GHG from the production of the operating energy and water

used during the life of the building happen every year. This graph reveals that over the life of a

building, the impact of the operating energy, dominated by heating and cooling, is predominant in

the whole carbon footprint of the building.

0

500

1000

1500

2000

2500

3000

3500

4000

4500

GW

P [

kgC

O2

-eq

/Occ

/yr]

MODULAR

TRADITIONAL

Figure 1: GWP over the whole life of Stella B17 Modular vs Traditional equivalent



To better visualise the differences in construction between modular and traditional, we have re-

moved from the graph the operating energy and water, as shown in Figure 2 below:

Looking at all the materials and services involved in fabricating, transporting and erecting these

buildings, the modular building has a carbon footprint 6% more than it’s equivalent in traditional

construction. To better understand this situation, the next graph Figure 3 shows the same compar-

ison but this time focusing on the elements of the buildings that are different between the two

construction methods:

In this graph, we have removed elements similar to the two buidings and we see that steel frame,

plywood and bathroom structure; all three specific to the modular construction, account for 30% of

these selected materials and services. The modules’ structure is made of steel frame and steel

sheets for the floors (aggregated under “Steel frame” in the graphs). Moreover, this modular build-

ing uses quite a lot of concrete for the ground floor and the two cores. Consequently, the carbon

footprint of the structural elements of the modular building is higher that the concrete used in the

traditional equivalent.

In a second order comes the additional carbon footprint of the double-gazing and the insulation in

the walls, but these elements participate to reducing the operating energy which, as we saw, is

more important than the carbon footprint of construction materials. Finally, the transport of the

modules, from Melbourne to Perth for that particular project, also add to the overall deficit of the

modular building.

-200

20406080

100120

GW

P [

kgC

O2

-eq

/Occ

/yr]

Traditional Modular

Figure 2 Stella B17 Modular vs Traditional - Carbon Footprint of Selected Material and Services

020406080

100120

Serv

. Eq

uip

Fitt

ings

…

Do

ors

&…

Car

pe

t

Ste

el f

ram

e

Wal

l fin

ish

Co

ncr

ete

…

Ce

ilin

g

Ply

wo

od

Co

ncr

ete

…

Tran

spo

rt o

f…

Pe

op

le &

Eq

uip

m

Insu

lati

on

(w

all)

Bat

hro

om

fin

ish

Bat

hro

om

…

Insu

lati

on

(sl

abs)

Flo

or

Tile

s

Ro

of

cove

rin

g

Bri

cks

Mo

rtar

Form

wo

rks

+…

GW

P [

kgC

O2

-eq

/Occ

/yr]

Traditional Modular

Figure 1 Stella B17 Modular vs Traditional - Carbon Footprint of Material and Services



For this particular project and without changing drastically the original design, one of the low hang-

ing fruits would be to replace carbon intensive materials with a mainstream low carbon equivalent.

For example, replacing the carpets with a timber floor.

4.2 The detached houses

4.2.1 Heating and Cooling loads

The 22 modular houses were modelled, along with their next neighbouring buildings to simulate

the shadow, to calculate the heating and cooling loads for Perth weather.

The best and worst performing modular houses, number 21 and 12, were selected to be compared to their

traditional equivalent (after modelling them) as shown in

Table 1 below.

Table 1 Thermal Performance of best and worst modular houses vs traditional equivalents

Construction House

#

HEATING

[kWh/m2]

COOLING

[kWh/m2]

TOTAL

[kWh/m2]

Modular 21 35.57 4.18 39.75

Traditional 21 61.63 1.73 63.36

Modular 12 42.06 5.44 47.50

Traditional 12 65.53 1.87 67.40

The added insulation of the modular houses (in the walls) helps contain the heat better in the winter and

requires less energy to keep the house at 21C. From

Table 1 we see that house 21 modular requires 42% less energy in winter than its equivalent in traditional

construction and house 12 modular 36% less than traditional. Both constructions benefit from a concrete

floor slab and have similar low inertia ceilings, but the traditional house has a heavier thermal inertia thanks

to its brick walls (external and internal) compared to the light steel-frame construction of modular. This

added inertia helps retain the coolness of the night in summer, hence requiring less energy to keep the house

at 24C. From

Table 1 we see that house 21 traditional requires 59% and house 12 traditional 56% less energy

in summer than its equivalent in modular construction. Though this lower performance in summer

has a minimal impact on the overall annual energy load.

4.2.2 Carbon footprint

Figure 4 below shows the carbon footprint, expressed as Global Warming Potential (GWP) meas-

ured in kgCO2eq/occ/y, of the modular house #21 compared to its traditional equivalent:

Figure 3 BGC house 21 Modular vs Traditional - Global Warming Potential



As for the multi-storey building, the carbon footprint of operational energy is dominant in the life

cycle assessment of the building.

Again the operating energy and water are removed from the graph Figure 5 in order to better

visualise the differences in construction between modular and traditional:

Looking at all the materials and services involved in fabricating, transporting and erecting these

two houses, the modular building has a carbon footprint 26% less important than it’s equivalent in

traditional construction. The next graph Figure 6 shows the same comparison but this time focus-

ing on the elements of the houses that are different between the two construction methods:

In this graph, we have removed elements similar between the two houses. We observe that the

mineral-based construction elements of the traditional house, namely the concrete, the roof tiles,

the bricks and the mortar, account for the major part of the difference between modular and tradi-

tional constructions. These are carbon intensive materials mainly due to their energy intensive

fabrication process. A focus on the roof shows that despite the fact that the timber structure of the

traditional house is three time less carbon intensive than the steel-frame of the modular house, the

weight of the terra cotta roof tiles makes this carbon intensive element predominant in the roof

category.

0.0020.0040.0060.0080.00

100.00120.00140.00160.00180.00200.00

Traditional

Modular

Figure 5 Comparison of GWP (in kgCO2eq/occ/y ) of selected materials and

services of traditional and modular BGC houses

Figure 4 BGC house 21Modular vs Traditional - Carbon Footprint of Material and Services



Additionally, the People & Equipment category is 29% less carbon intensive for modular. That is

because modular benefits from an industrialised fabrication process involving less staff and less

equipment. For example, the prefab concrete slab takes less staff than the traditional slab with on-

the-spot custom-made formworks. Steel roof structures are easier to assemble in a factory than

timber structures on-site, and the same can be said for the steel structure of the walls compared

to the on-site erection of traditional walls brick by brick or the use of plasterboards versus hand

laid plaster for finish. But more importantly, the fabrication and installation process of modular

being 126 days shorter than for traditional, that is 126 round trips the staff doesn’t have to make to

work on this one house, hence less carbon emissions from their vehicles.

For this particular project and without changing drastically the original design, the modular house

could improve its carbon footprint even more by adopting a prefabricated roof structure made of

timber trusses.

5. Conclusion and lessons learned

Other studies have looked at different factors of sustainability between modular and traditional

construction, such as construction waste, embodied energy, construction time and cost. These

two case studies comparing two similar buildings of two different construction methods teach us

some lessons from the thermal performance and carbon footprint perspective.

On the thermal performance:

1. If the lighter construction of modular buildings has many advantages, it also represents in

general a weak point for the thermal performance in summer.

2. More specifically, this weakness is particularly impacting the high-rise building typology

rather than the low-rise house typology, because of the higher window area and com-

pactness. Indeed, multi-storey buildings are less shaded by eaves and surrounding built

or natural environment. Moreover, compared to a house, average units tend to have less

surfaces in contact with the exterior (heating load) and still a fair amount of glazing (cool-

ing load).

3. But thermal inertia is beneficial to summer thermal performance only when night temper-

atures drop low enough below day temperatures. The thermal performance results of the

multi-residential building in the 3 different climate zones showed us that there were sig-

nificant differences. The lesson to draw is consideration of the local context is necessary.

4. That same case study also showed that solar gain is a key to easy thermal performance

improvement and should be considered right from design stage to limit cost and ensure

elegant integration of solar design.

On Carbon footprint:

1. Operating energy is a predominant element of the carbon footprint of a building in its life

cycle. Unless the building is near passive house standard (i.e. close to zero operating

energy from heating and cooling), then operating energy is the low hanging fruit to re-

duce Global Warming Potential of buildings over their life time.

2. Consideration of the carbon intensity of materials should be given right from the design

stage to allow for exploration of environmentally friendly alternatives at minimum, if not

zero, extra cost.

These lessons are the result of the study of these 2 particular projects and further research is

necessary to refine and generalise this knowledge.

The author believes that off-site construction can offer a real advantage over on-site construction

when it comes to carbon footprint and operating energy; while still retaining the other more out-

spoken benefits (waste, time …).

The manufacturing process of off-site construction being more repetitive and controlled than on-

site construction, it naturally welcomes continuous improvement. Therefore, off-site construction



appears to be the key to achieve the evolution required from the building sector to play its role in

reducing the national level of GreenHouse Gas emissions in line with the target set recently in

Paris (COP21).

The Hickory group has embraced continuous improvement from the start, both for their manufac-

turing process and their products. They learned very quickly from the Stella B17 experience and

the next generation of modules was already designed to achieve better summer performance and

use less carbon intensive materials.

To take this topic further, the author has engaged in a doctoral research to identify and define the

links between off-site construction and operating energy.

6. References

Aye, Lu, T. Ngo, R. H. Crawford, R. Gammampila, and P. Mendis. 2012. "Life Cycle Green-house

Gas Emissions and Energy Analysis of Prefabricated Reusable Building Modules." Energy

and Buildings 47: 159-168. doi: http://dx.doi.org/10.1016/j.enbuild.2011.11.049.

Byrne J, Eon C & Newman P 2014. "Josh’s House –Delivery of a Mainstream Stream Zero Emis-

sion Home." World Renewable Energy Council’s Annual Renewable Energy Journal.

Krug, Daniela. 2013. "Offsite Construction: Sustainability Characteristics." BuildOffSite.

http://www.buildoffsite.com/content/uploads/2015/03/BoS_offsiteconstruction_1307091.pdf.

Lewis, Miles. 2006. "Prefabrication in the Gold-Rush Era: California, Australia, and the Pacific."

APT Bulletin 37 (1): 7-16.

Monahan, J., and J. C. Powell. 2011. "An Embodied Carbon and Energy Analysis of Mod-ern

Methods of Construction in Housing: A Case Study Using a Lifecycle Assess-ment Frame-

work." Energy & Buildings 43 (1): 179-188. doi: 10.1016/j.enbuild.2010.09.005.

United Nations Environment Programme. 2009. "Buildings and Climate Change: Sum-mary for

Decision-Makers." Sustainable buildings and climate initiative.

US Environmental Protection Agency, Defining Life Cycle Assessment (LCA), 17 October 2010,

http://www.gdrc.org/uem/lca/lca-define.html

Cao, Xinying, Xiaodong Li, Yimin Zhu, and Zhihui Zhang. 2015. "A Comparative Study of Envi-

ronmental Performance between Prefabricated and Traditional Residential Buildings in

China." Journal of Cleaner Production 109: 131-143. doi:

http://dx.doi.org/10.1016/j.jclepro.2015.04.120.

Kim, Doyoon. 2008. "Preliminary Life Cycle Analysis of Modular and Conventioinal Housing in

Benton Harbor, Michigan." University of Michigan Ann Arbor.

http://www.gdrc.org/uem/lca/lca-define.html

http://dx.doi.org/10.1016/j.jclepro.2015.04.120



7. Annexes

Annexe 2 3D Model for Thermal Performance Simulation

Annexe 1 Architect view of Hickory's Stella B17 building



Annexe 3 Lay out of the 22 modular houses

Annexe 4 Street view of a BGC modular house



Annexe 5 Model of a modular house for Thermal Performance Simulation

Documents

Environmental sustainability of prefabricated modular res ...energy embodied in the building itself (Monahan and Powell in 2001), and some have shown that operating energy is the most